Refyner is the data tool built for eCommerce and subscription brands. Prep and schedule your everyday dataflows on Snowflake – with the models you run on, from CAC and LTV to repurchase and churn, ready out of the box.
Under two minutes: pull in a Snowflake table, prep and shape it with drag-in components, preview the data and the generated SQL, then schedule it – no engineering ticket, no second tool.
Every model other tools make you build from scratch, Refyner ships ready-built – a proven set of dataflows refined across dozens of live brands. Push one to your Snowflake in one click, then run it as-is or adapt it.
Most eCommerce and subscription data teams on Snowflake feel all three of these – and the first is where the other two come from.
Every routine join, clean-up or new model goes through engineering. You sit in a ticket queue for work you could do yourself – and so does the business waiting on you.
A prep tool, a separate scheduler, and the engineering time to wire them together and keep them alive. The bill climbs every renewal.
Requests pile up and the setup slows you down. You're shipping in sprints when the business is asking for answers in days.
The value isn't generic SQL – it's Refyner's understanding of your schema and the business logic in every model. An analyst describes the work; Refyner drafts it, checks it against your live warehouse, and watches it run – so what ships is correct, scheduled, and yours to control.
Pre-flight checks & run monitor
BR_SHOPIFY_ORDERS, DIM_CUSTOMERcustomer_id per cohort weekDIM_CUSTOMER isn't unique on customer_id – flagged in background review.Before a dataflow runs, Refyner validates it against your live Snowflake – every column, the grain, and any join that could fan out. After it ships, it reviews hand-built dataflows in the background and watches every scheduled run, so a broken job never becomes a wrong number on a dashboard.
Describe what you want in plain English. Refyner reads your Snowflake schema, picks the right tables, joins and grain, and drafts the dataflow for your analyst to review and adjust – falling back to a reference model or asking a clarifying question when the request is underspecified.
Generate
What this dataflow does
Creates weekly cohorts of repeat customers with an LTV over £500. Joins Shopify orders to the customer dimension, filters to the last 12 months, and refreshes every Monday at 06:00.
Point Refyner at any dataflow – one you built or one it generated – and it explains, in human terms, exactly what the thing does and when it runs. Analysts validate intent at a glance; stakeholders finally understand what's behind the number.
The difference: your models built in, run by an analyst, in one place, on your own Snowflake.
| Refyner | Power BI Dataflow | Alteryx | dbt / Airflow | |
|---|---|---|---|---|
| eCommerce & subscription models built in | Yes | Build it yourself | Build it yourself | Build it yourself |
| Built & run by analysts | Yes | Partly | Desktop only | Engineer-run |
| Prep and schedule in one place | Yes | Refresh only | Server to schedule | Two tools |
| Runs on your Snowflake | Pushdown | Fabric capacity | Extract or In-DB | Yes |
| Supported, with an SLA | Included tier | Microsoft support | Paid tiers | DIY |
See the detail: vs Alteryx · vs Power BI Dataflow · vs dbt / Airflow
"CX Data gave us a working data tool overnight – and the models were live in weeks."
"We get board-ready profit updates every morning – and never hired a data engineer to get them."
Your raw data stays in your Snowflake either way. Choose how Refyner runs – see the security model.
The fastest way to start. Connect your Snowflake and build – we host and run Refyner, you get every update automatically.
Refyner runs inside your own environment – nothing leaves your boundary. Pair it with a support package and we'll help you run it.
Refyner is a data prep and scheduling tool for eCommerce and subscription brands. Analysts build and schedule dataflows on their own Snowflake, with pre-built models – LTV, blended CAC, retention, RFM, MRR movement and churn – built in. Refyner is a CXDATA product.
Heads of data and analysts at eCommerce and subscription (D2C) brands running on Snowflake who want to ship everyday dataflows themselves, without waiting on data engineering.
No. Refyner runs as pushdown on your own Snowflake – transformations execute in your warehouse and your data stays there. Refyner's control plane only holds metadata.
dbt and Airflow are two separate, engineer-run, code-first tools. Refyner puts prep and scheduling in one analyst-run platform with eCommerce and subscription models built in – and it's supported.
Refyner Cloud starts at $99/month for 5 users. Data prep, dataflow building and sampling are free; scheduled dataflows on Snowflake are charged at $3 per compute hour. A self-hosted deployment is available for data-residency needs.
A model is a set of dataflows that produces an output for a business subject area – for example Customer LTV, Blended CAC, Retention & repurchase, RFM segmentation, MRR movement, and Churn & cohorts.
Connect your Snowflake and your analysts prep and schedule it themselves – models built in, no engineering ticket, no second tool.
